On the Discriminant Validity of Burnout, Depression and Anxiety: A Re-examination of the Burnout Measure

We investigated the discriminant validity of the frequently used Burnout Measure (BM) (Pines et al ., 1981), in relation to depressive symptomatology and anxiety. Following Russell's (1980) work on the multidimensional space of affective states, BM's items were hypothesized to reside in three separate regions of that space, labeled anxiety, depression, and wornout, with wornout representing a combination of physical and emotional exhaustion. Respondents were 704 senior army officers, who completed a questionnaire that included BM items, standard measures of depression and anxiety, and measures of military-related stresses. Their responses to the BM items and to the depression and anxiety scales were subjected to multidimensional scaling analysis. Excepting an unexpected finding related to a fourth construct, loss of vigor, the results were supportive of the hypothesis and led to the construction of three unidimensional measures, respectively gauging anxiety, depression, and wornout. The correlations of the three measures with the stress measures supported their discriminant validity. We suggest that future research using the BM should focus on the set of items that were found to load on the dimension of wornout, to the exclusion of items tapping depression and anxiety.